OtherJun 11, 2026, 05:52 AM
Researchers say they trained a foundation model from scratch for about $1,500
Summary
Researchers at Sapient developed HRM-Text, using a Hierarchical Recurrent Model architecture, training a 1B-parameter foundation model from scratch for about $1,500 with only instruction-response pairs. The model achieves competitive performance with much larger open models, dramatically reducing pretraining costs and enabling organizations with limited resources to train their own reasoning models.
Why it matters
This event demonstrates a new method for low-cost foundation model training, potentially shifting the AI industry's reliance on massive compute and data.
Source links
⚠ Content is from official & reputable-media public sources, AI-assisted and auto-published, for information only — not investment advice.
Market reaction
The following is market reference information for related companies, and does not constitute investment advice.
Samsung Electronics
KRX · 005930